A study out of the University of British Columbia in Canada that was recently published wanted to determine the validity and reliability of the Fitbit One — a popular health wearable.

As the physical activity wearable market is exploding and millions of devices are being utilized, it’s important to see the true capabilities and limitations of these devices.

For those of us in academia, these studies bring reassurance to the notion of prescribing activity trackers to our patients — especially those that have been validated.

The study by Takacs, Pollock, and colleagues1 had the following methods:

  • 30 participants without any MSK pathology (so basically, healthy patients).
  • Participants walked on a treadmill at five different preset speeds — video was used to record the steps, and then two observers counted the steps independently.
  • Participants walked for 30 minutes.
  • 3 Fitbit One’s were placed on the participant: one on each hip side, and then the front pocket of pants or shorts (per Fitbit’s claims on how to use the wearable).
  • Distance traveled determined by treadmill output.

When it comes to measuring steps, Fitbit did very well. The percent relative error was below 1.3% for all tested treadmill walking speeds. The study noted that several other fitness trackers that have attempted validation prior have had issues being accurate at slower speeds, and clearly Fitbit’s algorithm did not have any issues with this.

fitbit accuracy

That said — Fitbit did not do well when it came to measuring distance. At slower speeds, Fitbit would over estimate distances walked. For example, at a speed of 0.90(m/s), Fitbit was measuring 0.28 kilometers traveled, versus the treadmill output of just 0.20 kilometers.  Obviously, over time, that error in distance will only grow. The authors mentioned how this could give inaccurate information to physicians and patients since those with chronic diseases such as COPD do not have brisk gaits.

The distance measured on the Fitbit uses a default step length based on height and gender, possibly explaining this distance inaccuracy, and the authors mentioned there are ways to change this within the Fitbit settings.

Another finding of the study was the flexibility of placement of the device. The device was placed in three different locations on the body, and all locations yielded similar results.

Overall, this study is reassuring for those of us interested in prescribing health wearables to our patients. Even though the Fitbit One is more than a year old, this study not only showed the validity of step counts, but it showed how there has been a tremendous improvement in the accuracy of wearables over the past few years.  The study authors mentioned several past wearables that have been attempted to be validated in the past that did not have accurate measurements.

How to prescribe to patients: 

Physician’s should consider prescribing the Fitbit One to their patients, but caution them their device should not be used solely for distance traveled, but for steps performed. They should tell their patient’s with chronic diseases that the step count will still be accurate for them, but their distance traveled will be more skewed if they have a slower walking speed.  Further, it’s important to note the study results were based on healthy individuals with no MSK pathology.  Patient’s who have MSK and neurological pathologies will not have as accurate results.